[ HRS.DEV // PORTFOLIO ]
[ HENDRY RAMA SETHIAWAN ]
[ SEMARANG / ID · UTC+7 ]
[ OPERATOR: HRS ]
HRS.DEV
// SYSTEMS ARCHITECT · AI/ML SPECIALIST · FULL-STACK DEV
[
]
- LOADING 0% -
Skip to Main Content
H
HRS.DEV
ID
EN
// OPERATOR FILE — HENDRY RAMA SETHIAWAN

Hendry Rama Sethiawan

Systems Architect · AI / ML Specialist · Full-Stack Developer

Systems Architect holding a Master's degree in Electrical Engineering (AI & Machine Learning focus). Architecting high-performance web systems and resilient digital ecosystems. Creator of jalaran.com AI workspace.

DEGREE / FOCUS
M.Eng. // AI & ML
FLAGSHIP
JALARAN.COM
CORE STACK
NEXT.JS / FASTAPI
Hendry Rama Sethiawan
SEMARANG / ID · UTC+7
HRS.DEV
M.ENG. ELEC. ENG // AI & ML
AVAILABLE FOR HIRE
CLASS
SYSTEMS ARCHITECT
EDUCATION
M.T. ELEC. ENG
PRIMARY DOMAIN
HRSETHIAWAN.COM
STATUS
ACTIVE / DEPLOYABLE
// 01

Codex

OPERATOR_PROFILE.DAT
3D MATRIX // BOUNDING CUBE
LIVE 3D ROTATION
AXONOMETRIC LATENT CUBE
MOAT :: ACTIVE 3D
Systems Architect AI / ML Bridge EE & Signals FinTech / Advisory
// BASE LOCATION
Semarang / Malang, Indonesia [IDN] · UTC+7
// ACADEMIC STANDING
M.Eng. in Electrical Engineering (GPA 3.94)
B.Eng. Cum Laude (GPA 3.69)
// LOCAL TIME
--:--:-- WIB
// DOSSIER BIOGRAPHY

Systems Architect & AI/ML Researcher with a Master of Engineering in Electrical Engineering & Bachelor of Engineering (Cum Laude) from Universitas Brawijaya, specializing in Machine Learning for Anomaly Detection and Real-Time Data Pipelines.

Experienced in engineering real-time sensor data pipelines at BRIN (National Research & Innovation Agency), ensemble anomaly detection thesis research (Isolation Forest & LOF), and building jalaran.com AI productivity workspace using Next.js (App Router) & FastAPI.

// CORE COMPETENCIES
Unsupervised Anomaly Detection
Isolation Forest · LOF · Robust Mahalanobis · PCA
M.Eng. thesis — 4-channel FSR array, 2,400 samples, 0.948 ROC-AUC
Quantitative Capital Markets
GARCH(1,1) · TimeSeriesSplit · scikit-learn · yfinance
IDX Quant Screener & Market Primer — 49 tickers, walk-forward validation
Real-Time Data Pipelines
UDP sockets · pandas · numpy · real-time validation
BRIN research internship — 8-channel sensor acquisition over UDP
Full-Stack Architecture
Next.js · FastAPI · Laravel · Redis · PostgreSQL
Jalaran — modular monolith, queued ML, locally-served AI
THESIS RESEARCH
Ensemble Anomaly Detection
Ensemble Isolation Forest & LOF model with scikit-learn for anomaly detection.
BRIN EXPERIENCE
UDP Sensor Data Logger
Real-time health device pipeline collecting 8-sensor streams.
FLAGSHIP PLATFORM
Jalaran Workspace
Unified AI productivity workspace with 3-panel shell.
// 02

Operations

RECORD.ARCHIVE
Jalaran unified AI productivity workspace, three-panel interface
[ FLAGSHIP WORKSPACE ]
OP.01 / JALARAN STATUS: PRODUCTION READY
AI WORKSPACE MODULAR MONOLITH

JALARAN WORKSPACE

Unified AI-assisted productivity workspace bringing AI chat, guided ML data analysis, RSS news reader, Kanban task management, note-taking, and file processing tools together behind one consistent interface. Powered by a local privacy-preserving AI layer and a 3-panel workspace shell.

Next.js (App Router) TypeScript FastAPI Python scikit-learn Redis Queue PostgreSQL
Ensemble anomaly detection benchmark summary for a four-channel FSR sensor array
#02
2025 · TESIS M.T. UNSUPERVISED ML

ENSEMBLE ANOMALY DETECTION

Unsupervised fault detection on a 4-channel force-sensitive-resistor array sampled through a 12-bit ADC. Three detectors with deliberately different inductive biases — Isolation Forest, LOF, and Robust Mahalanobis (MCD) — fused into one calibrated score. Benchmarked against ground truth: across 2,400 samples with 288 injected faults, the median ensemble reached 0.948 ROC-AUC and 0.979 precision at a 10% alert budget.

Python scikit-learn Isolation Forest LOF Robust Mahalanobis PCA
IDX Quant Screener diagnostic dashboard showing ranked undervaluation scores
#03
2025 · QUANTITATIVE CAPITAL MARKETS

IDX QUANT SCREENER

A quantitative value screener for the Indonesia Stock Exchange. Pulls fundamentals and daily prices for 49 IDX tickers plus the ^JKSE benchmark, folds them into a single undervaluation score, and ranks the universe. A LogisticRegression + StandardScaler model is validated walk-forward with TimeSeriesSplit, emitting a CSV shortlist and a diagnostic chart.

Python scikit-learn yfinance TimeSeriesSplit pydantic-settings loguru
IDX Market Primer report panels: volatility, drawdown, and correlation matrix
#04
2025 · QUANTITATIVE VOLATILITY MODELLING

IDX MARKET PRIMER

A quantitative starter kit for the Indonesian market. Pulls daily price history from Yahoo Finance, computes return, volatility and maximum-drawdown metrics, fits a GARCH(1,1) model per ticker, builds a correlation matrix, and scores data quality before writing the results out as reports.

Python GARCH(1,1) yfinance pandas numpy
Sethry Business AI Consultant chat interface
#05
2025 · PERSONAL LLM & FLASK

SETHRY BUSINESS AI CONSULTANT

A business consulting assistant for Indonesian MSMEs. A Flask web application running the Qwen3-0.6B language model locally through Hugging Face Transformers, with automatic CUDA/CPU device selection, category-specific prompting (finance, marketing, operations, HR), batch processing, rate limiting, and a health-check endpoint.

Flask Transformers Qwen3-0.6B PyTorch Rate Limiting
UDP sensor data logger architecture diagram
#06
BRIN BANDUNG · 2022 SENSOR PIPELINE

UDP SENSOR DATA LOGGER

A real-time acquisition pipeline for wearable health monitoring devices, built during a research internship at BRIN Bandung. Listens for UDP datagrams across 8 sensor channels, timestamps each packet, validates NaN/inf values inline, and accumulates them into a pandas DataFrame for CSV export.

Python UDP Sockets pandas BRIN Research
// 03

Service Record

HISTORY.LOG
// SUMMARY
06
Years in active service
Years Active 06
GPA (M.T.) 3.94
National Funding 01
Research Projects 3+
Competitions 03
// CITATION & FUNDING

"Delivered a superior machine learning model in thesis research (GPA 3.94) and secured Rp 5.68 million national funding as PKM-K Team Leader."

— Universitas Brawijaya & PKM-K Evaluation
AUG 2025 — NOV 2026 (EXPECTED)

M.Eng. in Electrical Engineering Candidate (GPA 3.94 / 4.00) · Universitas Brawijaya

Specializing in unsupervised machine learning for anomaly detection. Thesis: Anomaly Detection in Plantar Pressure Data Using Isolation Forest, LOF, and Combined Model. Advanced coursework in Computational Intelligence, Advanced Programming, and Stochastic Systems.

MAY 2022 — JULY 2022

Research Intern · National Research and Innovation Agency (BRIN), Bandung

Engineered real-time sensor data pipeline for continuous wearable health monitoring. Collected thousands of data points across 8 sensors using Python UDP sockets, pandas DataFrames, and real-time NaN/inf validation.

JAN 2022 — MARCH 2023

Deputy Head of Department · Brawijaya Aeronautics Workshop

Revitalized post-pandemic workshop operations, managing 30+ team members and executing participation in 3 national engineering competitions while boosting team retention and coordination.

DEC 2020 — JAN 2023

Laboratory Assistant · Digital Systems Laboratory, Universitas Brawijaya

Guided electronic logic comprehension for 50+ engineering students per semester, managed practical lab curriculum, and supervised 9 assistant team members.

DEC 2020 — SEPT 2021

PKM-K Team Leader · Student Creativity Program in Entrepreneurship

Secured Rp 5.68 million in national funding for custom Android electronics repair application project, leading a 5-member cross-functional engineering team.

AUG 2019 — JAN 2023

B.Eng. in Electrical Engineering (GPA 3.69 / 4.00 Cum Laude) · Universitas Brawijaya

Completed Bachelor's degree with Cum Laude honors. Final Project: Health Monitoring System for Patients via User Datagram Protocol (UDP) and LoRa wireless radio.

// 05

Recruitment

CONTACT.CHN
recruit@hrsethiawan:~ CONTACT
$ initiate --recruit
> Establishing secure channel...
> Handshake complete.
> Channel encrypted.
> Awaiting transmission.
$
// NAME REQUIRED
// INVALID EMAIL
What do you need?
// MESSAGE REQUIRED

// GOES STRAIGHT TO MY INBOX — NO EMAIL CLIENT NEEDED

// STATUS
Available for deployment

Open for architecture advisory, AI/ML pipeline integration, and high-performance web development engagements. Response time under 24 hours for all inquiries.

// CHANNELS
// NOTICE

For NDAs, source code review, or architecture briefs, reach out via the primary contact channel.